OBJECTIVE To determine the relationship between brachial blood pressure, and transfer function-estimated and invasively measured central aortic pressure in patients with at least moderate symptomatic aortic stenosis. METHODS Fourteen patients aged 54-81 years with mean (SD) effective valve area of 0.69 (0.20) cm2, undergoing coronary angiography, had simultaneous peripheral and central aortic blood pressure measurements. Brachial blood pressure was determined by an oscillometric method. Aortic pressure was measured directly using pressure transducer tipped catheters, and estimated indirectly by the application of a transfer function to a radial arterial waveform obtained by tonometry. RESULTS Measured aortic systolic pressure did not differ significantly from brachial pressure [mean difference (SD) 2 (9) mmHg, P = not significant (NS)]. Transfer function estimates of central systolic pressure obtained from the radial waveform calibrated from brachial pressure were less accurate [mean difference -8 (7) mmHg, P = 0.001]. Recalibration of the radial waveforms using the invasive mean and diastolic blood pressure improved the agreement [mean difference -2 (6) mmHg, P = NS] but did not provide a better estimate than brachial blood pressure. The accuracy of noninvasively estimated subendocardial viability ratio was substantially improved by recalibration of radial arterial waveforms using corrected ejection time. CONCLUSIONS In patients with aortic stenosis there is clinically acceptable agreement between noninvasive brachial pressure and directly measured central aortic pressure.